HANDBOOK:QUALITY PRACTICES INBASIC BIOMEDICAL RESEARCH

(1)

Special Programme for Research and Training in Tropical Diseases (TDR) sponsored by UNICEF, UNDP, World Bank and WHO

HANDBOOK :

QUALITY PRACTICES

IN BASIC BIOMEDICAL RESEARCH

Prepared for TDR

by the Scientific Working Group on Quality Practices

in Basic Biomedical Research

(2)

ACKNOWLEDGEMENT

This handbook is designed to address the very important topic of quality in basic biomedical research. The quality practices outlined in this handbook provide the basis for a non-regulatory quality management system, which if applied properly, will enable institutions and individuals to produce credible research data. The quality management system is designed as an aid to research institutions and individual researchers wishing to improve the quality of their research data.

The existing draft Handbook on quality standards in basic biomedical research (2001).

was reviewed by a specialist working group, convened by UNICEF/UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases (TDR), consisting of scientists from around the world. An editorial group was mandated to write the current document based on comments from the working group. The revisions and additions suggested by the working group have been integrated into this current, expanded Handbook on quality practices in basic biomedical research (QPBR)..

The quality management system outlined in this Handbook will provide institutions and researchers with the necessary tools for the implementation and monitoring of quality practices in their research, thus promoting the credibility and acceptability of their work. The handbook highlights non-regulatory practices that can be easily insti- tutionalised with very little extra expense.

TDR gratefully acknowledges the participation and support of all those involved in the production of this Handbook. Special thanks go to the editorial group, Nadya Gawadi, David Long and Jürg Seiler for their dedication and immense contribution.

For all correspondence:

Dr Deborah Kioy PhD Pre-clinical Coordinator Product Development and Evaluation

Special Programme for Research and Training in Tropical Diseases World Health Organization

20 Avenue Appia 1211 Geneva 27 - Switzerland

Tel. +41 22 791 3524 Fax.+41 22 791 4774 E-mail: [email protected]

p-tit.qxd 16.2.2006 11:54 Page 2

(3)

FOREWORD . . . 7

SCOPE AND PRINCIPLES . . . 11

1. INTRODUCTION TO QUALITY PRACTICES IN BIOMEDICAL RESEARCH . . . 13

2. WHAT IS BASIC BIOMEDICAL RESEARCH ? . . . 17

2.1 Basic biomedical research : drug development model . . . . 17

2.1.1 Stage 1a : discovery per se . . . 18

2.1.2 Stage 1b : transitional research . . . 18

2.1.3 Stage 1c : non-regulated, non-clinical research . . . 18

2.2 Basic biomedical research other than drug development . . . 18

2.2.1 Stage 1a : discovery stage per se . . . 19

2.2.2 Stage 1b : transitional research . . . 19

2.2.3 Stage 1c : practical proof of principle (POP) stage . . . 19

3. WHAT IS QUALITY IN RESEARCH ? . . . 25

3.1 Relationship between study and data . . . 27

3.2 Data, records and report . . . 28

3.3 Reproducibility . . . 29

3.4 The purpose of quality practices . . . 29

4. THE QUALITY PRACTICES IN BASIC BIOMEDICAL RESEARCH . . . . 31

4.1 Organization . . . 31

3

FOREWORD

This document is the revised version of the draft handbook Quality standards in basic biomedical research (TDR / PRD / QSBR / 01.1), published in 2001. The draft handbook was based on the deliberations of a specialized scientific working group (SWG) convened by the UNDP / World Bank / World Health Organization Special Programme for Research and Training in Tropical Diseases (TDR).¹ The SWG, composed of inde- pendent scientific experts, met in Geneva 4-6 September 2000 to discuss various aspects related to Good Laboratory Practice (GLP) and sound scientific working practices. They concluded there was a pressing global need for dissemination of information and guidance to cover basic biomedical research.

To this end, the draft document – a preparatory guide outlining the quality standards required in disease endemic countries for basic biomedical research – was produced.

This preliminary publication was intended to serve as a working document, the first step towards developing an agreed set of guidelines on quality practices for basic biomedical research. It was circulated internationally to scientists and researchers involved in research on new strategies for preventing or fighting disease, including the development of new drugs. Comments and suggestions were canvassed with a view to incor- porating feedback into a revised, final document.

To prepare this final document, an editorial group was put together to collate all the comments received in response to the draft document and to prepare a draft of the final document. The document was then discussed by a group of experts from a wide range of scientific disciplines at a review meeting in Geneva, 25-26 January 2005. The objectives of the review meeting were to :

• review the draft final document for content, scope and coverage of all important aspects

• review the appendices for suitability and applicability

1 Now the UNICEF / UNDP / World Bank / WHO Special Programme for Research and Training in Tropi- cal Diseases

(8)

• make recommendations for the publication of a final document, which was given a new title : Quality practices in basic biomedical research.

The quality practices for basic biomedical research described in this document do not address the scientific content of a research programme or proposal, but are concerned with the way the research work is managed.

8

Foreword

p-tit.qxd 16.2.2006 11:54 Page 8

(9)

9

Foreword

Participants to the Review group Meeting, Geneva. 25 - 26 January 2005 Mr David Bailes, South Barn Consulting Ltd, United Kingdom

Dr Hervé Bossin, Food and Agriculture Organization (FAO) / International Ato- mic Energy Agency (IAEA) Agriculture and Biotechnology Laboratory, Austria Dr Gerard Daly, In Step Training, Ireland

Professor KS Gamaniel, National Institute for Pharmaceutical Research and Development, Nigeria

Dr Agneta Ganning, AstraZeneca R&D Södertälje, Sweden Dr Nadya Gawadi, Maxygen ApS, Denmark

Dr Myriam Arevalo Herrera, Instituto de Immunologia del Valle, Colombia Mr Paul A Lemetti, BioAid, Pennsylvania, USA

Mr David Long, CHIMEX, France

Dr MJ Moshi, Muhimbilli University, College of Health Sciences, Institute of Traditional Medicine, Tanzania

Dr Catherine Mundy, Center for Health Systems and Services, Management Sciences for Health, Massachusetts, USA

Dr Juerg Seiler, ToxiConSeil, Switzerland

Dr Marianne Martins de Araujo Stefani, Federal University of Goias, Instituto de Patologia Tropicale, Brazil

Dr Sudhir Srivastava, Central Drug Research Institute, India Professor Vincent Pryde K Titanji, University of Buea, Cameroon Professor A Walubo, University of the Orange Free State, South Africa Dr Dorcas Yole, Institute of Primate Research, Kenya

Dr Mariano Gustavo Zalis, University Federal do Rio de Janeiro, Instituto de Biofisica Carlos, Brazil.

(10)

10

Foreword

WHO Secretariat

Dr Hashim Ghalib, Research Capability Strengthening (RCS) / TDR

Dr Deborah Kioy, Preclinical Coordinator, Product Development and Evaluation (PDE) / TDR

Dr Janis Lazdins, Acting Coordinator PDE / TDR Dr Ali Mohammadi, PDE / TDR

Dr Ayo Oduola, Functional Coordinator Strategic and Discovery Research (SDR)/

TDR

Dr Rob Ridley, Director TDR

Dr Fabio Zicker, Functional Coordinator RCS / TDR.

Editorial Group

Dr Nadya Gawadi, Maxygen ApS, Denmark Mr David Long, CHIMEX, France

Dr Nina Mattock, SSK / TDR

Dr Deborah Kioy, Preclinical Coordinator, PDE / TDR Dr Juerg Seiler, ToxiConSeil, Switzerland.

p-tit.qxd 16.2.2006 11:54 Page 10

(11)

Basic biomedical research refers to the use of fundamental scientific principles in medical and biological research directed towards developing tools to detect, prevent or treat human disease. Basic biomedical research is commonly encountered in the discovery and exploratory stages of product / drug development. This type of research is not yet covered by any agreed national or international regulations and guidelines. Since production of credible research data can only be ensured by compliance with sound study management, the proposed ‘quality practices’ are concerned with the organizational, managerial and practical aspects of basic biomedical research. This handbook is a com- panion volume to the TDR handbook on quality practices for regulated safety research.¹ This current handbook offers guidance and tools for the practical management of basic biomedical studies, ensuring that ideas are translated into action, and that ensuing information is accurately captured in order to enter the public domain.

The handbook :

• Defines basic biomedical research in the context of drug development and existing regulations and quality systems.

• Examines the organizational and practical framework for basic biomedical research studies irrespective of type (drug development or other fields of health research).

• Examines the role of prescriptive and descriptive documentation.

• Discusses how to record, report, review, archive and publish results in order to bring them into the public domain.

• Covers ethical concerns and biosafety.

SCOPE AND PRINCIPLES

1 TDR. Handbook : Good Laboratory Practice. Quality practices for regulated non-clinical research and dev- elopment.Geneva, UNDP / World Bank / WHO Special Programme for Research and Training in Tropical Diseases (TDR), 2001 (TDR / PRD / GLP / 01.2).

11

Chapter 1 • Introduction to quality practices in biomedical research

(12)

• Provides templates for standard operating procedures (SOPs), curricula vitae and training records, and some sample SOPs.

It is hoped that wide application of the quality practices proposed in this handbook will lead to cost-effective, accelerated discovery research and will ultimately benefit human health.

12

Scope and principles

p-tit.qxd 16.2.2006 11:54 Page 12

(13)

The world’s population is facing serious health challenges in the form of newly emerging diseases or disease patterns e.g. avian influenza, severe acute respiratory syndrome (SARS), transmissible spongiform encephalopathies (bovine spongiform encephalopathy, Creutzfeldt-Jakob disease), human immunodeficiency virus (HIV), Ebola, and multidrug resistant diseases or organisms such as malaria. There are in- creasing difficulties in treating ‘old’ diseases such as trypanosomiasis, onchocerciasis, diabetes, hypertension and cancer. The problem is worsened by the changing age distribution in populations, greater population movements that promote transmission of diseases, new practices in land use, agriculture and forestry, and changing world cli- mate, to name but a few. As a result, there is increased demand for new drugs and new principles for treatment, based on new knowledge about the causes and mechanisms of diseases, and for new methods of vector control. The search for these commodities and principles increases the need for scientific researchers and research programmes. With the continued restrictions in available funding, it is essential that basic scientific research as a whole, and especially in all fields connected with health issues, be conducted in a proper fashion using processes that minimize waste of resources and reduce the need for costly confirmation and repetition of work already performed.

Today, research facilities in many universities, hospitals, government institutions and industries are used for basic scientific studies relevant to the discovery and development of new strategies for fighting disease including products with potential usefulness in health care. Data from these activities need to be reliable to ensure a solid basis for deciding whether to invest in further development of a strategy or product. Since the activities fall outside regulatory scope, i.e. they are not covered by, for example, the Organisation for Economic Co-operation and Development (OECD) Principles of Good Laboratory Practice (GLP), a need for guidance on quality practices in these areas has been recognized. This is why this handbook was commissioned.

It should not be surprising, therefore, to find that some controversies in the scientific literature could probably have been resolved earlier, more easily and better if the practical experimental conditions had been fully described, or if the supportive data had been properly collected.

13

1. INTRODUCTION TO QUALITY PRACTICES

IN BIOMEDICAL RESEARCH

(14)

It must be stressed here that the quality practices for biomedical research described in this document do not address the scientific content of a research programme or proposal, but are concerned with the way the research work is organized and planned, performed, recorded, reported, archived, monitored and published. Figure 1 sketches the main steps in this process.

14

Figure 1. Flow of research activities

DOCUMENTED PROCESSES MONITORED PROCESSES

PLAN

DO

CAPTURE

REPORT

STORE &

ARCHIVE

PUBLISH

A high level proposal describes the research programme Narrower detailed plans define the individual studies

A range of activities ensue

Various controls should prevent artifact

The outcome of the activities must be recorded as raw data

The results are reported

The results are moved into the public domain

Usually biomedical research is first outlined in a research proposal and then described more fully in individual study plans or protocols explaining why and how the experimental work will be undertaken. It is clear that if the basic underlying conditions of the experimental set-up are unclear or poorly documented, and if the raw data are incomplete, there may be fundamental doubts about the validity of the knowledge obtained and its contribution to science. The application and use of sound scientific principles in the conduct of basic exploratory and discovery studies, coupled with an attention to good quality practices, will optimize transparency in data and reporting.

The availability of clear evidence will facilitate the choice of new strategies for pre-

chap.1.qxd 16.2.2006 11:56 Page 14

(15)

15

venting or fighting disease, including the choice of new drug candidates for further development. Not only will investment in the wrong candidate waste resources and time, but it will also block capacity that could have been used for developing a more promising candidate.

(16)

chap.1.qxd 16.2.2006 11:56 Page 16

(17)

Basic biomedical research refers to activities intended to find means of detecting, preventing or treating human disease. Such research covers the discovery and exploratory studies that precede the regulated phases¹ of drug development or programmes to develop other methods of disease control. These later stages of biomedical research are time consuming and require vast financial, human and technical resources ; this is true whether new basic principles are being established or new drug candidates developed.

2.1 Basic biomedical research : drug development model

The candidates for further development come from the first R&D stage in basic biomedical research, often called ‘discovery’. This discovery stage can be further subdi- vided into three : discovery per se, transitional research, and non-regulated non-clinical research (see figure 2). In table 1, the contents of each stage are illustrated by reference to three examples.

Figure 2. The three stages of basic biomedical research in drug discovery

17

1 The regulated stages are :

• the non-clinical stage to establish drug safety

• the clinical stage to establish safety and efficacy in man

• the post-approval stage where the drug is monitored for safety, and its production carefully controlled.

For a brief account of basic biomedical research in the context of the regulated drug development process, and the quality practices appropriate to regulated stages, see Appendix 2.

2. WHAT IS BASIC

BIOMEDICAL RESEARCH ?

Stage 1a Discovery per se

Stage 1b

Transitional research

Stage 1c Non-regulated Non-clinical research

(18)

2.1.1 Stage 1a : discovery per se

During this stage, the researcher identifies a possible new drug candidate. To begin with, the researcher notices signs that a compound may have therapeutic potential and finds ways to establish whether or not it is a fruitful lead to follow. The idea for the compound may come from direct observation, scientific literature, knowledge of traditional practices, or systematic screening. It is unlikely that the research progresses smoothly ; the researcher will probably meet many dead ends and collect many inconclusive results. The researcher may not be able to formulate a testable hypothesis or make a firm plan (see table 1).

2.1.2 Stage 1b : transitional research

During this stage, the researcher tries to characterize the active pharmaceutical ingre- dient (API) and starts to investigate how to produce and analyse it, while continuing focused biological experimentation to investigate its actions in cells, tissues or the whole body (see table 1).

2.1.3 Stage 1c : non-regulated, non-clinical research

During this stage, biological tests on subcellular systems, tissues and / or whole animals provide evidence for efficacy – i.e. ‘proof of principle’(POP). These are rigorously controlled studies with biological models. They are needed to indicate whether the compound is biologically active, and whether it is likely to be efficacious in man, before time and resources are invested in the formal, expensive and regulated stages of drug development. A sufficient supply of well-characterized test compound has to be ensured (see table 1).

Thus this stage provides non-clinical proof of principle (not to be confused with non- clinical safety or clinical studies – see appendix 2).

2.2 Basic biomedical research other than drug development

Even if the research does not concern drug discovery and development, the stages are analogous, as illustrated in figure 3. In the discovery stage, ideas are formulated and tested by observation or experimentation. The contents of each stage of basic biomedical research for products other than drugs are illustrated by three examples in table 2.

18

Chapter 2 • What is basic biomedical research ?

chap.2.qxd 16.2.2006 10:35 Page 18

(19)

Figure 3. The three stages of basic

biomedical research for products other than drugs

2.2.1 Stage 1a : discovery stage per se

During this stage, observations verify that there is a phenomenon worth pursuing.

2.2.2 Stage 1b : transitional research

Having studied the available literature, the researcher attempts to formulate the causal relationships in his conceptual model of the situation in order to prepare the relevant experiments for stage 1c.

2.2.3 Stage 1c : practical proof of principle (POP) stage

Here the researcher demonstrates, in the experimental model, that the right relationships have indeed been identified. The researcher may also include the first studies to show the potential applicability of these insights as practical methods for disease control, prevention or cure. However, here the practical proof of principle stage does not include animal or field studies to demonstrate safety, nor any clinical studies that involve dosing or other interventions in man – these belong to the later stages (see appendix 2).

Table 2 provides examples of stages 1a, 1b and 1c.

19

Stage 1a

Discovery stage per se

Stage 1b

Transitional research

Stage 1c Practical proof of principle research

(20)

Table 1. Examples illustrating the three stages in basic biomedical research for drug candidates

20

Example A A researcher suspects that one of the body’s own macromolecules might be exploited to alleviate a progressive, crippling condition, common throughout the world. In the discovery stage the researcher makes efforts to establish whether this is a reasonable idea, and how the molecule might function. He / she finds out how to obtain a supply of the molecule for the next stages of research.

Example B A researcher knows that a population traditionally uses a local herb to alleviate an affective disorder. But the herb contains dozens of interesting compounds, of which several might be the active principle. The researcher needs to establish whether only the herb is used, or whether other methods or drugs are used simultaneously. The herb is identified, and its habitat mapped. It is very difficult to establish at this stage whether the herb is effective or whether the population is enjoying a placebo effect. Although an observational study might be relevant at this stage, it is important to emphasize that this is not the right stage for a clinical study.

Example C A large company with expertise in psychiatric disorders routinely screens many thousands of potential compounds from a ‘drug library’ shared with a neighbouring company which has expertise in another field. The screening programme is designed to detect molecules with a good ‘fit’ to the target receptors, and dozens are found each week. A database is used to list and profile the promising candidates, and further cheap screening tests are used to narrow the field further.

Stage 1a

chap.2.qxd 16.2.2006 10:35 Page 20

(21)

21

A universal problem arises : the body’s own molecules quickly enter the body’s own metabolism and do not remain at therapeutic levels long enough to be effective. The researcher might reason that if the molecule is modified by adding side chains, the extra bulkiness might slow its breakdown or elimination, extending the period of therapeutic activity. Activities at this transitional stage include finding the optimal number and type of side chains, and optimizing methods for production and analysis. The researcher experiments with possible formulations. There are also biological tests to establish binding properties and other characteristic.

Kinetic studies in intact animals provide evidence that the modified molecules stay longer in the body. Receptor binding studies might show that the molecule targets the right cells or tissues. An animal model may be used to test potential efficacy.

Having isolated the most promising active pharmaceutical compounds (API), the researcher further explores the biological activity in cell, tissue and / or animal models. He / she investigates whether to extract the compounds or whether to produce them chemically or by biotechnological methods. Further research includes optimizing methods for producing and analysing the compounds, as in the previous example.

Receptor binding studies and animal behav- ioural models are most useful for establishing potential for efficacy.

Having identified the most promising active pharmaceutical compounds (API) from the extended screening, the company explores the biological activity in cell, tissue and / or animal models. The company investigates how to produce the compound in quantities suitable for further testing. Further research includes optimizing methods to analyse the compounds, as in the previous example.

Receptor binding studies and animal behav- ioural models are most useful for establishing potential for efficacy.

Stage 1c Stage 1b

(22)

22

Table 2. Examples illustrating the three stages in basic biomedical research for products other than drugs

Example X The researcher has noticed a relationship between the health of a population and their water supply, in comparison with a neighbouring population that enjoys better health. The researcher verifies, in an epidemiological study, that the health status really is different from that of the neighbouring group.

Example Y A community needs to find the cause of an unusual pattern of neurological disease.²The prevalence of the neurological problem is mapped, verifying that the prevalence is higher than expected.

Example Z A researcher suspects that the behaviour of an invertebrate vector is ge- netically controlled, and that it is possible to reduce the mating success of the vector by changing the genome .³The researcher attempts to map the genomes of the vector and a similar invertebrate with different behaviour.

Stage 1a

2 Sacks O. The island of the colour blind.Book II : Cycad Island.New York / Toronto, Alfred A. Knopf, 1997.

3 Demir E, Dickson BJ. fruitless splicing specifies male courtship behavior in Drosophila. Cell, 2005, 121(5) :785-794.

chap.2.qxd 16.2.2006 10:35 Page 22

(23)

23

Is the problem really the water ? The researcher decides what parameters to measure, and how to demonstrate causality. The researcher considers practical solutions to the problem.

The researcher demonstrates, using the experimental model, that the causal relationship has been identified, and perhaps includes the first studies to indicate the applicability of this insight as a practical method for disease prevention.

The researcher decides what to examine in the way of causal factors, e.g. diet, heredity, pollu- tion. How will the assays be set up, and how will the suspected culprit be tested ?

The researcher demonstrates, in an experimental model, that the right factor has been identified, and perhaps includes the first studies to show how to avoid or control the problem.

The researcher considers how to identify the relevant genes, and how to associate them with the mating behaviour.

The researcher demonstrates, in an experimental model, that the right genes have been identified, and perhaps includes the first studies to show that modified mating behaviour can be used as a practical method of disease control.

Stage 1c Stage 1b

(24)

chap.2.qxd 16.2.2006 11:58 Page 24

(25)

There are two aspects of quality in research : a fundamental scientific aspect, and a practical experimental aspect. When the underlying science is wrong or the working hypothesis is ill conceived, the results obtained by even the best conducted experiments will not lead to a true advance of knowledge. On the other hand, even the best science, the most brilliantly reasoned working hypothesis, will not return results and answers that are acceptable to the scientific community if they are not supported by flawlessly conducted (i.e. high quality) experiments.

The matrix below explains how the quality of research and the quality of science are interrelated.

25

3. WHAT IS QUALITY IN RESEARCH ?

Table 3. How sound scientific principles and good quality practices contribute to the credibility of results

Sound scientific Good quality Credibility of

principles practices results

Scientific study 1 No No No

Scientific study 2 No Yes No

Scientific study 3 Yes No No

Scientific study 4 Yes Yes Yes

It is often claimed that science is self-policing in that research proposals and results are exposed to the scrutiny of peers, and are challenged by attempts to repeat and verify the experiments. However, the extent to which this process ensures the integrity of research and its results has been questioned. Broad and Wade¹described many instances where spurious results were published and survived for some time. They traced the anomalies to organizational practices within the scientific endeavour itself. But also outright

1 Broad W, Wade N. Betrayers of the truth. Fraud and deceit in science.Oxford University Press, 1982.

(26)

fraudulent machinations and manipulation of data may not be ruled out, as recent expe- rience and some highly publicized cases have shown. A group of scientists from the American Institute of Medicine²therefore recommended closer surveillance of scientific activities in order to assure reliability of new scientific results, and a Danish group³ came to similar conclusions. Both groups called for clear organization of work, clear definition and allocation of responsibilities, close supervision of scientific work, good data recording practices, good facilities for data storage and retrieval, and proper training of scientists and other staff.

This handbook is intended to guide scientists in how to organize their research and add value to it by promoting the credibility of their data. It should be possible for peers, scientific journals, development partners, or authorities to audit studies to verify authenticity and reliable reporting of results, and this would help to validate the data and make the results acceptable to the scientific community at large, making it more likely there will be commensurate returns on the investment. The scientific community also needs to be able to repeat studies in order to confirm the knowledge and build further research on the results.

Normally quality is defined as “the totality of characteristics of an entity … that bear on its ability to satisfy stated and implied needs”.⁴ It is clear that there is a need to obtain research results that are solid enough to enable development of useful products and principles for fighting diseases. In order to fulfil these needs, scientific research must deliver results that are :

• Relevant

• Reliable and reproducible

• Ethical

• Auditable

• In the public domain.

These guidelines endeavour to put practices in place to ensure that results with these characteristics are achieved.

26

Chapter 3 • What is quality in research ?

2 Institute of Medicine. Report of a study. The responsible conduct of research in the health sciences.Wash- ington DC, National Academy Press, 1989.

3 Andersen D et al.Scientific dishonesty and good scientific practice.Danish Medical Research Council, 1992.

4 ISO 8402 : 1994 : Quality management and quality assurance – vocabulary. International Standard 2^ndedition, 1994.

chap.3.qxd 16.2.2006 10:37 Page 26

(27)

3.1 Relationship between study and data

Scientific activity must generate reliable data. Figure 4 shows the relationship between plan, experiment, data and interpretation.

27

Figure 4. Study plan, study activities, data, interpretation (arrows denote variables likely to influence data integrity).

Study plan

Data

Interpretation Study

activities

Meaningful scientific interpretation of study results is only possible when founded upon reliable data. Clearly, in order to obtain reliable data, the experimental variables that always affect studies must be kept under control. Quality practices are designed to help scientists control the variables. Such an approach is the only way to obtain reliable results and sound scientific interpretation, and to avoid being dogged by ‘false positives’

and ‘false negatives’. This reasoning explains why best practices attach so much im- portance to precise planning (through the written study plan or protocol) and to use of standardized techniques (by following previously written standard operating procedures).

Where there is an experimental set-up, the scientist must put effort into planning and setting up controls or control groups. If the experiment involves exposing animals or in vitro systems to demonstrate pharmacological activity, a vehicle control must be

(28)

included as a minimum. The inclusion of a positive control is recommended as proof that the set-up is working. If the experiment includes analytical assays, the samples must include blanks and known quality control (QC) standards.

All sources of bias in the experimental or assay set-up should be explored and, if possible, brought under control. All influences and inputs to the study must be considered before the activities begin.

If the study is a purely observational study of events as they happen, without any intervention, information on the physical or social setting is vital for understanding the results. It is important to plan which parameters to study, to ensure a systematic and continuous record. Needed are : date, time, precise place, identity of the collector, the study ID, and the conditions which make each observation relevant and valid. For example, if you are comparing the behaviour of insects in different locations, you have to know whether you are comparing the same or different species before collecting the main body of data, and you need to define what constitutes the interesting differences in habitat. Or, if you wish to compare a group of people who visited a practitioner of traditional medicine with a group visiting a modern hospital, you have to check thor- oughly the therapeutic practices at each location before starting observations.

In all cases, it is wise to obtain the advice of a statistician in order to collect enough data to be able to compare the results between groups and draw conclusions. The statistician will also advise on choice of appropriate statistical tests for the amount and type of data anticipated.

3.2 Data, records and report

The account of the experimental surroundings, set-up and preparations is as vital to the understanding of results as are the results themselves. The study plan must include methods for limiting bias and sources of error. The records must contain descriptions of the whole experimental set-up, or the conditions for collecting specimens, or the conditions for the observational study. The data must be fully identified with respect to time, place, study identification and collector, and be followed and checked throughout processing and presentation. Finally, the report must contain descriptions of these aspects.

28

chap.3.qxd 16.2.2006 10:37 Page 28

(29)

3.3 Reproducibility

Reproducibility is one way of testing the reliability of data. This means that if the investigator or someone else were to repeat the experiment in the same set-up, equivalent data would result. Or if specimens were collected in the field, another visit to the same or a similar habitat at the same time of day / year would yield a similar collection. Since especially significant, valuable or controversial findings often require confirmation by repeating a study, all studies should be designed, managed, controlled, recorded and reported sufficiently to ensure reproducibility of the findings.

3.4 The purpose of quality practices

The practices outlined below are intended to increase the likelihood that – provided the research has a scientific basis and the hypothesis is testable – research activities will generate reliable data suitable for publication and perhaps for further research aimed at detecting, preventing or treating disease. The use of quality practices should reduce the risk of obtaining inconclusive results on account of uncertainty about controls or because of unclear procedures. The use of quality practices should also change attitudes to certain aspects of research management that are not widespread today : routine supervision, review and audit, as used to confirm authenticity and veracity of results.

29

(30)

chap.3.qxd 16.2.2006 12:03 Page 30

(31)

4.1 Organization

4.1.1 Quality policy and staff responsibility

It is essential for each research institution to have a written policy statement describing the quality practices to be applied by all personnel in the conduct of experimental work, irrespective of its nature. This statement need not be lengthy – it could for example just refer to this document. The policy statement would be supported by written guidelines outlining responsibilities at the different organizational levels. The administration (director) of the organization should be visibly and fully supportive of these measures, and should implement mechanisms for their application, exercising at least some level of control over them.

Responsibilities should include at least those in table 4.

31

4. THE QUALITY PRACTICES

IN BASIC BIOMEDICAL RESEARCH

(32)

Table 4. Matrix of responsibilities, roles and activities in basic biomedical research

32

Chapter 4 • The quality practices in basic biomedical research

1 The intention is to present the minimum roles necessary for implementing the TDR quality practices.

Actual jobs and positions, e.g. student, researcher and lecturer at different levels, would be mapped into these roles for the purpose of allocating responsibilities. For example, it would be possible for someone with the rank of professor, or with the rank of PhD student, to take the role of principal scientist. The principal scientist may in reality also be the head of department. In a very small team, one person may be filling the roles of head of institute, departmental leader, principal scientist and technician. The principal scientist plays the key role in ensuring study quality, in the sense that a study cannot be performed without a principal scientist.

2 This role is becoming more widespread at research institutions, but is by no means universal or com- pulsory. Quality assurance personnel take an active role in implementing and maintaining quality measures, and, through their auditing activities, keep management informed of the level of compliance to quality requirements.

Responsibility for scientific activities

Organizational role¹ Responsibility

Director of organization Policy, provision of resources of all types, budget, supervision of activities.

Responsibility for review of scientific activities

Ethics

Quality assurance (QA) personnel²(if relevant)

Assist in implementation and maintenance of quality practices. Help to assure the authenticity, traceability, and consistency of data, and compliance with WHO / TDR quality practices.

Head of department Use of resources, supervision, advice and support for junior staff, compliance with institutional policy and WHO practices.

Principal scientist (sometimes called principal investigator (see appendix 1))

Conduct of study, scientific interpretation of study results, veracity of study data.

Technician Performance of procedures as required in study plan and SOPs, or other instruction.

All personnel Compliance with the ethics manual.

Management Ethics committee.

Ethics manual for the institution.

Other support staff Fulfil duties according to instruction.

Peer Scientific analysis and collegial criticism.

chap.4.qxd 16.2.2006 10:38 Page 32

(33)

This model can be adapted to suit the needs of a small team or an individual researcher – see appendix 10.

33

4.1.2 Personnel and training

The organization’s administration should ensure that the responsibilities of staff at all levels are defined and documented in job descriptions. Aspects to be considered are : scientific field of activity, practical duties, supervisory duties / delegation, administrative and financial responsibilities, communication, and keeping knowledge and skills up to date. There must be sufficient financial support for education and training activities.

While it is important to ensure that all staff are knowledgeable and well versed in the relevant quality aspects of experimental work, including planning, recording and reporting, this is especially relevant for tutors, PhD students and postdoctoral fellows in university settings. These staff members should be responsible for full application of the quality practices required by the institution and should not deviate from these. This requirement also holds for any institution where new staff are employed on a tempo- rary, project-related basis.

Qualifications and training should be adequate for the activities that each person is to carry out. At the time of recruitment, management should verify the authenticity of the qualifications documented in the curriculum vitae (CV) by, for example, contacting named referees or checking on publications. While at the institution, staff may become further qualified through a structured course of study and may be awarded a diploma or degree from a recognized academic institution. These qualifications must be suitable for the type of research envisaged and should be documented in the person’s CV, be verifiable and kept up to date.

Training is necessary for all staff (at all levels) to prepare them for using specialized or new techniques. Such training should be offered routinely ; it will help to keep the

Quality policy and staff responsibility

• The research institution should establish a written policy describing its quality practices.

• The responsibilities of each level of personnel should be defined and documented.

KEYNOTE

(34)

general level of expertise at the research institution up to date. Remember that training should be completed before practical activities start. This training should also be documented in separate training records held by the institution.

34

4.2 Physical resources

The research institution’s administration is responsible for providing facilities of suitable size, construction and location, and for providing suitable equipment to meet the requirements of the research programme and its individual studies. Fulfilling the requirements of the study does not necessarily mean that state-of-the-art constructions or equipment have to be provided. Instead, the institution’s administration, in coopera- tion with the leader of the research group or the principal scientist, must carefully con- sider the objectives of the research programme, including its individual component studies, and decide how to achieve these with the facilities and equipment available in the local environment. Special attention should be paid to any risks for study integrity that could originate from the close proximity of different activities and studies ; care should be taken to minimize any potential interference with the validity of the study, especially regarding the risk of confusion and mix-ups (of studies, test systems, test items, data) or cross contamination (for example, between chemical compounds or strains of micro-organisms). The purpose of such requirements is to ensure that the individual study is not compromised because of inadequate facilities or equipment.

All equipment should be suitable for its intended use both in the laboratory and the field. The actual choice of equipment will be based upon the scientific requirements for accuracy, precision, robustness and measurement interval. This choice is a scientific task and will be made by the institution's scientific management. It is good practice to draw up user requirement documents or specifications for equipment before purchase.

Personnel and training

• All personnel should have written job descriptions.

• All study staff must keep their CVs up to date.

• Training records of all staff members should be kept up to date.

KEYNOTE

chap.4.qxd 16.2.2006 10:38 Page 34

(35)

Once acquired, equipment must be properly calibrated and maintained, if necessary by a qualified and certified agent, to ensure accurate and consistent performance and to ensure that measurements are comparable to those from other laboratories. Calibration of balances is a good place to start (because one can follow up by calibrating volumetric equipment). Typically, the institution has a set of standard weights with which to cali- brate the balances – at least once a year (secondary calibration). If there is an accred- ited standardization organization in the country, representatives check the institution’s standard weights once every three years (primary calibration). In addition, daily checks with one standard weight will ensure validity of the weighing. Other equipment is calibrated in a similar way. It should be noted that equipment, adequately calibrated in laboratory conditions, may no longer be calibrated when transported into field conditions, and will routinely need rechecking before use.

Records of repairs, routine maintenance and any non-routine work should be retained. This is to ensure the reliability of data generated and to prevent loss or cor- ruption of data as a result of inaccurate, inadequate or faulty equipment.

35

Physical resources

• Facilities must provide adequate protection in order to avoid putting the studies at risk through mix-ups, confusion, or cross contamination.

• Equipment must be suitable for use in the study; suitability should be supported by documentation.

• A calibration and maintenance programme for equipment must be established, documented and maintained.

KEYNOTE

4.3 Documentation

Making a full record of all information is essential not only to permit appropriate scientific interpretation of the results but also to enable complete reconstruction of the study, should this be necessary. Documentation is the only way of demonstrating what actually went on at the time of the experiment. Without documentation the process is meaningless ; essentially there has been no study.As well as containing the data generated,

(36)

the study records must prove that all the required procedures were correctly carried out at the stipulated time. If complete records are not made, the study validity is compromised. Missing data suggest that either the procedure concerned was never performed or that the data have been lost. In either case the study is seriously compromised and may have to be repeated from scratch. Documentation may be divided into two broad classes :

• Prescriptive documentsthat give instructions as to what is to happen during the course of a study.

• Descriptive records that describe what actually happened during the course of the study.

Figure 5 illustrates the relationship of prescriptive and descriptive documents to one another and to the study activities.

Figure 5. Prescriptive documents, study activities and descriptive documents

36

QPBR SOPs

Data

Publication Report DESCRIPTIVE DOCUMENTS

PRESCRIPTIVE DOCUMENTS

Study plan Study

activities

Prescriptive documents include research proposals, study plans and standard operating procedures, and are a preparation for practical activities. Descriptive documents include raw data, any derived data, study reports and publications. For example, the

chap.4.qxd 16.2.2006 10:38 Page 36

(37)

dosing instructions in a study plan are prescriptive – they do not constitute proof that the animals were dosed. One has to make specific records to show that this was the case – these are descriptive.

Appendix 9 shows another way of presenting the hierarchy of prescriptive and descriptive documents, setting them in context with training as the means of translating written instruction into managed activity.

37

General documentation

• Research institutions should maintain both ‘prescriptive’ and ‘descriptive’ documents.

• Research institutions should ensure that there are full records of all study activities, sufficient to provide complete study reconstruction.

KEYNOTE

4.3.1 Prescriptive documents : research proposals and study plans

In basic research, a study, or a set of studies, is usually outlined as part of a research proposal. A research proposal is a document outlining 1) the scientific context, 2) the overall objectives, and 3) the scope (or thrust) of a research programme. Although the proponent, usually a research scientist(s), is responsible for the scientific content of the proposal, all research scientist(s) responsible for running the programme must be indicated in the proposal. The proposal should also outline the main stages in the research or, in the case of a large programme, describe the individual component studies and indicate the general timeframe of each study in the overall programme. Normally a review board or the management within the institution approves the document, since human and financial resources have to be found to support the work. It is essential that each research institution has a policy or guidelines describing the generation, review and approval of research proposals. Most grant-giving organizations provide an application form to guide the contents and layout of the proposal.

The study plan (or study protocol) is a document describing in detail the proposed conduct of an individual study. Because it is the key document for communicating the intentions of the study to all contributing staff and sponsors, its contents and layout should be clear. Every study must therefore have a plan which is in line with the overall objectives of the research proposal.

(38)

For clarity, the research institution should define in writing the relationship between the research proposal and the study plan(s), and the responsibilities attached to each (see Fig. 6).

Figure 6. Relationship between research proposal and study plans

38

Study plan 1 Research

proposal

Study plan 2

Study plan 3

The study plan should describe the study design in detail, including the purpose, intended methods, and names of persons who will carry out the study and interpret the experimental data. Proposed dates for key events should also be stated.

The information in the study plan should be sufficiently detailed to enable the study to be repeated exactly, if necessary. The study plan is, therefore, likely to contain details of the :

• test material and conditions for its handling and storage

• type and quality of reagents and equipment

• type of test system and how it is to be handled

• observations to be made

• methods for data collection, evaluation, verification and (if appropriate) statistical analysis

• methods for reporting and archiving results

chap.4.qxd 16.2.2006 13:58 Page 38

(39)

• ethical implications of the experiment, where appropriate (for example, in research involving animals), and discussion on this topic.

As noted above, procedures considered routine in the laboratory may be described in standard documents of the institution, such as SOPs. Therefore the study plan need not explain such procedures in full detail but could instead reference the relevant SOPs. It is good practice for the research institution to standardize the sections needed in all study plans.

If an experiment is based on previous preliminary work, this work would normally be referenced in the plan to ensure traceability to the early data, justifying certain parameters investigated in the main experiment. References cited in the study plan should either come from published peer reviewed sources or from internal research reports where data or documentation are available. Both published sources and internal reports must be verifiable. The link between the proposed activities and the published material must be explicit.

The principal scientist has the authority for final approval of the study plan. He / she signs to show his / her full responsibility for performing the study according to the plan and in compliance with quality practices in biomedical research (QPBR). The principal scientist should ensure that the technicians responsible for day-to-day conduct of the experimental stages are familiar with the study plan and its associated procedures. Such instruction should be documented in the study notes.

Subsequent major, intended, changes to the study plan will need authorization – in a document called the ‘study plan amendment’ – from the principal scientist since any change may entail significant modification to the scientific purpose of the study. How- ever, minor deviations from the plan may simply be recorded in the laboratory note- book or recorded on specifically designed data sheets which are then filed with the rest of the data.

39

(40)

Research proposal and study plan

• The research institution should define the difference between the research proposal and the study plan.

• The research institution should have guidelines for the production, review and approval of research proposals.

• The research institution should have guidelines for the production, review and approval of study plans.

• Each individual study should be the subject of a single detailed study plan (one study = one study plan).

• The research institution should provide a format and a list of minimum contents for a study plan in accordance with QPBR recommendations.

• The research institution must make it clear that the principal scientist’s signature on a study plan indicates that he / she takes full responsibility for the conduct of the study according to the plan and according to QPBR.

KEYNOTE

40

4.3.2 Standard operating procedures

Standard operating procedures (SOPs) are documents that provide instructions for activities of a repetitive, routine nature in a very detailed manner. Such standardized, approved, written working procedures are required by classical quality assurance techniques, indeed by good management.

Remember the quotation generally attributed to Dr Joseph M Juran :

“Use standards [i.e. SOPs] as the liberator that relegates the solved problems to the field of routine, leaving the creative faculties free for the problems that are still unsolved”.³ SOPs follow a defined life cycle: writing, approval, distribution, update, and withdrawal.

3 http://specializedqualitypublications.com/B.htm.

chap.4.qxd 16.2.2006 10:38 Page 40

(41)

Every institution, or laboratory or facility within an institution, will certainly already have a collection of standard procedures under various headings and in differing forms, e.g. recipes for the preparation of buffer solutions or tissue culture media, directions for the operation and maintenance of apparatus and instruments, or step-by-step instructions for commonly performed activities (appendices 4-6 provide examples). These should become integrated into a fully coherent system with a standard layout, which should be predefined by management. This should lead to centralized organization of formatting, numbering, issuance, modification, withdrawal and archiving, which will help avoid duplication of effort, incoherence, delays, lack of traceability and incomplete distribution. The system should thus encompass all standard activities and there should not be separate, conflicting systems for conveying directives to personnel, such as memos.

The greatest benefits of such a system of standardized procedures can be expected if there is comprehensive coverage of :

• Standard scientific techniques

• Equipment, disposables, and reagents

• All critical stages of study design, management, conduct and reporting

• ‘Scientific’ administrative policy and procedures (e.g. format, safety and hygiene, security, personnel management).

Ideally, the individuals most familiar with the activity to be described in an SOP should also write the document. Furthermore, there should be somebody responsible for each SOP (author or other person responsible) to handle queries and keep each procedure updated. It is a good idea to impose a minimum time requirement for periodic review.

It is perfectly acceptable to use a manual if this is more appropriate than an SOP for the purpose. In this case, the SOP would reference the manual by name, edition, and physical placement. The SOP would be reviewed like any other SOP, taking special account of the continued relevance of the manual in relation to the current equipment.

Once the equipment is retired, the SOP and manual must be withdrawn and archived.

SOPs should be immediately available to all individuals performing the respective tasks. Staff must fully understand the SOP and follow it rigorously. Deviations from the standard way of performing these activities should be handled like deviations from study plans, i.e. they have to be described, justified, signed and dated, in order to pre- serve the credibility of the system.

For reasons of traceability and ease of use, a two-tier system of SOPs is often the pre- ferred approach. In such an approach, one tier may reflect general policies and pro-

41

(42)

cedures (e.g. protocol writing, review, approval, distribution and modification, SOPs, general rules for equipment use and maintenance, archives). The second tier may repre- sent technical methods (e.g. histological staining methods, analytical methods, specific procedures for use and maintenance of equipment). It is advisable to present the SOPs (SOP manuals) as a binder with an up-to-date table of contents, logical chapter divi- sions and selective distribution, to avoid a mushrooming pile of dust-gathering paper that often gets misplaced. All alterations to SOPs must be made through formal revisions ; notes and changes in the form of handwritten comments in the margin are not admissible.

All withdrawn SOPs, whether no longer used or superseded by a revised version, should be archived carefully in order to make a complete historical record of the test facility’s procedures.

Properly designed SOPs will bring the following benefits to the laboratory :

• Standardized, consistent procedures (person-to-person and test-to-test variability minimized)

• An opportunity to optimize processes

• Capture of technical and administrative improvements

• Demonstration of management commitment to quality as part of the SOP approval process

• Ease of documenting complicated techniques in study protocols and reports (a simple reference to the procedure is often sufficient)

• Continuity in the event of personnel turnover

• Availability of a training manual

• A means of study reconstruction after the event, even after a lapse of years

• A means of communication in the event of audit, visits, technology transfer.

The successful implementation of SOPs requires :

• Sustained support from all levels of management with commitment to establishing SOPs as an essential element in the organization and culture of the laboratory.

• SOP-based education and training of personnel so that all personnel perform the procedures in the same way.

• An effective sound SOP management system to ensure that current SOPs are available in the right place.

42

chap.4.qxd 16.2.2006 10:38 Page 42

(43)

4.3.3 Descriptive documents : good record keeping

Data should be collected in strict compliance with the rules of the institution. It is important to identify who collected the data from a given procedure and when. This is why it is necessary to sign and date all data at the time of recording.

Raw data are defined as all original recordings made during the course of a study.

The data should indicate :

• What was done– demonstrating compliance with the study plan.

• How it was done – demonstrating compliance with practical, experimental instructions (in the study plan and relevant SOPs).

• When the work was performed– demonstrating the existence of the events and their sequence in time.

• Who did the work– demonstrating conformity with the responsibilities delegated by management to suitably qualified personnel.

Characteristics of the collection of good raw data are :

• Attributability – data can be traced to their source, e.g. by study number, sample number, parameter. Unique identification of data pertaining to an individual study helps to prevent mix-up of data.

• Originality – raw data constitute the first recording of the observation. Raw data should not be recorded on scraps of paper for transcription into a final form. If you use a computer to record or capture the data, you need to define whether the signed

43

Standard operating procedures (SOPs)

• Each research institution must establish appropriate SOPs covering the activities of the research institution and the studies performed.

• The contents of SOPs should follow a standard format set by the research institution.

• The institution must implement a system for the management of SOPs. This will cover the writing, signature, issuance, modification, withdrawal and archiving of SOPs.

• The institution provides and records SOP-based training.

KEYNOTE

(44)

printout or the electronic records constitute raw data. If you decide to define the electronic record as raw data, the computer must be protected by password, and be backed up frequently.

To ensure these characteristics, data must be recorded :

• Promptly – data have to be recorded immediately the operation is completed. It is not acceptable to make the record some time after the job has been finished, since memory may fail or become inaccurate, which may lead to data loss or faulty records.

• Accurately – the raw data have to be a true representation of the observation, and accuracy is thus absolutely central to the integrity of the study.

• Legibly – data that cannot be read are useless, and records that are difficult to deci- pher raise doubts in the minds of the reader as to their credibility. Therefore, write legibly.

• Indelibly – one of the common problems in research is that data are often recorded in pencil and are subject to subsequent changes without this being evident, which in turn may lead to suspicions of deliberate tampering. Use of indelible and waterproof ink eliminates this problem. Any changes to raw data should be made so as not to obscure the previous entry. The person responsible for the change (or the person approving it) should then sign and date the change, and the reason for the change should be indicated, if necessary. It is furthermore necessary to check the robustness of the print-out from instruments : some fade quickly at room temperature or when stored in plastic folders. In such cases, an authorized (signed and dated) photocopy should be prepared for storage.

Data should be recorded and organized in a way that supports and facilitates both recording and all subsequent processes (e.g. data entry, reporting, audit, archiving).

The total collection of records, including raw data, analyses, printouts, and any other documents relevant to the study, constitute the study file. This study file also contains the study plan and amendments, and the study report. The study file, together with all documents relating to the study, are stored in the archives (see also section 4.3.6.).

Figure 7 shows how the study file is related to the research proposal and study plans.

44

chap.4.qxd 16.2.2006 10:38 Page 44

(45)

4.3.4 Use of notebooks

During the early discovery stage, some organizations require researchers to use notebooks to record all activities in the laboratory or the field. Usually these notebooks are numbered consecutively, sometimes by the laboratory administration, sometimes by the

45

Figure 7. Relationship between research proposal, study plans and study files

Study plan 1 Research

proposal

Study file 1

Study plan 2 Study file 2

Study plan 3 Study file 3

Good record keeping

• Each research institution must implement rules regarding the recording of raw data.

• Raw data and other records should be sufficiently detailed and complete to ensure study traceability and reconstruction.

• If computers are used to acquire, modify, manipulate or archive data, the raw data must be clearly defined.

KEYNOTE

HANDBOOK:QUALITY PRACTICES INBASIC BIOMEDICAL RESEARCH

Special Programme for Research and Training in Tropical Diseases (TDR) sponsored by UNICEF, UNDP, World Bank and WHO

HANDBOOK :

QUALITY PRACTICES

IN BASIC BIOMEDICAL RESEARCH

Prepared for TDR

by the Scientific Working Group on Quality Practices

in Basic Biomedical Research

ACKNOWLEDGEMENT

TABLE OF CONTENTS

FOREWORD

SCOPE AND PRINCIPLES

1. INTRODUCTION TO QUALITY PRACTICES

IN BIOMEDICAL RESEARCH

2.1 Basic biomedical research : drug development model

2. WHAT IS BASIC

BIOMEDICAL RESEARCH ?

2.2 Basic biomedical research other than drug development

3. WHAT IS QUALITY IN RESEARCH ?

3.1 Relationship between study and data

3.2 Data, records and report

3.3 Reproducibility

3.4 The purpose of quality practices

4.1 Organization

4. THE QUALITY PRACTICES

IN BASIC BIOMEDICAL RESEARCH

4.2 Physical resources

4.3 Documentation